Piotr Smaron

Results 207 comments of Piotr Smaron

@kbr-scylla pinging, since you requested for changes.

CI reports an error during build stage: ``` 19:04:31 cql3/statements/alter_keyspace_statement.cc:51:37: error: call to 'format' is ambiguous 19:04:31 51 | on_internal_error(mylogger, format("get_abs_rf_diff expects integer arguments, " 19:04:31 | ^~~~~~ ``` Interesting,...

UPDATE: rebased onto the latest master to fix the build.

This test `test_tablets.test_read_of_pending_replica_during_migration[true].dev.2` failed on adding a node to the cluster - https://jenkins.scylladb.com/job/scylla-master/job/scylla-ci/11852/testReport/junit/(root)/test_tablets/test_read_of_pending_replica_during_migration_true__dev_2_8/, and failed only for the dev mode, which is a strong indicator it's flaky. I'll restart the...

> This test `test_tablets.test_read_of_pending_replica_during_migration[true].dev.2` failed on adding a node to the cluster - [jenkins.scylladb.com/job/scylla-master/job/scylla-ci/11852/testReport/junit/(root)/test_tablets/test_read_of_pending_replica_during_migration_true__dev_2_8](https://jenkins.scylladb.com/job/scylla-master/job/scylla-ci/11852/testReport/junit/(root)/test_tablets/test_read_of_pending_replica_during_migration_true__dev_2_8/), and failed only for the dev mode, which is a strong indicator it's flaky. I'll restart...

@kbr-scylla @avikivity the CI is green after rebasing, can be queued again.

I'm guessing it'd be good to finalize these: - [ ] https://github.com/scylladb/scylladb/issues/17740 - [ ] https://github.com/scylladb/scylladb/issues/17904 - [ ] https://github.com/scylladb/scylladb/issues/19153 While the rest can probably be deferred.

OK, this was because I was using: ``` $ cqlsh --version cqlsh 5.0.1 $ which cqlsh /snap/bin/cqlsh ``` When using "our" `cqlsh` I see the error message correctly reported: ```...

> (not sure why "ConfigurationException") Maybe it's because of differences between different scylla versions this CQL query is tested against or maybe it's because you use yet another `cqlsh`?